This competing continuation application has three overall goals: First, to extend the reach of vaccination, by increasing the number of effective epitopes. Second, to understand and manipulate epitope dominance and subdominance, permitting us to maximize vaccine-induced immunity. Third, to clarify the effects of serial microbial infections upon vaccine-induced immunological memory. These three goals are divided into five specific aims: Aim 1. To identify subdominant epitopes and to carefully map the kinetics of the CTL responses to all epitopes, dominant and subdominant, during acute infection and in the memory phase. Aim 2. To extend the reach of vaccination by expanding the number of virus sequences open to immune surveillance. Can we convert a very poor epitope into a strong one? Aim 3. To design vaccines which circumvent immunodominance, and to evaluate the protective benefits. Current vaccines are limited by immunodominance; they induce responses only to the few dominant epitopes they contain. Can we design new vaccines which circumvent this problem, permitting the concurrent induction of responses to all encoded epitopes? Aim 4. To determine the role of lysis of antigen presenting cells in control of immunodominance. Is the phenomenon of immunodominance controlled by lysis of APCs by the "strongest" CTL responses? Aim 5. How stable is vaccine-induced memory for CTL, and what are the effects of subsequent microbial infections? If memory is diminished by subsequent infection, as has been suggested, this could have profound implications.